GPU Server
과학 연산, 데이터 분석, 머신러닝 추론, 비디오 트크랜스코딩과 같은 난이도 높은 연산 처리가 부각되면서 데이터를 신속/정확하게 산출하여 다변화된 비즈니스 환경에 대응이 가능한 컴퓨팅 자원으로 2개의 GPU부터 10개 GPU까지 병렬 서버로 연결하는 최고의 자원을 선택할 수 있습니다.


GNR4-A4DA8L2U8G


| AMD EPYC™ 9004 dual-processor 4U GPU server that supports eight dual-slot GPUs

| up to 24 DIMM, 11 PCIe 5.0 slots, Dual NVMe, four 3000W Titanium power supplies

| OCP 3.0 and ASMB11-iKVM


  • ● Powerful performance : Powered by AMD EPYC™ 9004 processors with 128 Zen 4c cores, 12-channel, up to 4800 MHz DDR5 and support for a maximum TDP of up to 400 watts per socket
  • ● AI and HPC workloads ready : Up to eight dual-slot active or passive GPUs, NVIDIA NVLink® bridge, and NVIDIA Bluefield DPU support to enable performance scaling
  • ● Power-efficient system design : Independent CPU and GPU airflow tunnels for thermal optimization and support for up to four 3000W Titanium redundant power supplies for uninterrupted operation
  • ● Cooling solutions : Enhanced air cooling based on CPU TDP for versatile workloads
  • ● Scale-up storage and expansion design : A total of eight bays in combination of Tri-Mode NVMe/SATA/SAS drives on the front panel and 11 PCIe 5.0 slots for higher bandwidth and system upgrade
  • ● Flexible networking module design : Optional OCP 3.0 module with PCIe 5.0 slot in rear panel for faster connectivity
  • ● Enhanced IT-infrastructure management : ASUS ASMB11-iKVM remote ontrol with ASPEED AST2600, ASUS Control Center IT management software and hardware-level Root-of-Trust solution

Accelerate performance for HPC and AI workloads


Eight dual-slot GPU compute in 4U

AMD EPYC™ 9004 dual-socket server built for the demands of enterprise AI infrastructure to deliver unprecedented performance with industry-leading GPUs, faster GPU interconnect and higher-bandwidth fabric, and supports up to eight dual-slot active or passive GPUs scalable configuration with a choice of NVIDIA NVLink® Bridge or AMD Infinity Fabric™ Link to enable scaling of performance with increased bandwidth — ready to match your AI and HPC workloads.

PCIe 5.0 Ready

Faster storage, graphics and networking capabilities

PCI Express® (PCIe®) 5.0 delivers 32 GT/s bandwidth, which is twice the speed of PCIe 4.0, and offers lower power consumption, better lane scalability and backwards compatibility. ASUS servers are PCIe 5.0 ready, with scalable designs to satisfy the increasing workloads or modern data centers.


System Layout

Pannel Layout

Block Diagram

MLPerf Benchmarks


Top results on AI workloads

ASUS NVIDIA-Certified GPU servers have achieved top performance results in the data-center Closed division across six tasks on MLCommons® MLPerf™ training and inference benchmarks, outperforming other servers with the equivalent GPU configurations to create complete, optimized solutions and cultivate strong industry partnerships to enhance AI developments in diverse fields to push technology to its limits.


Optimized Thermal Design

Independent CPU and GPU airflow tunnels

Our servers features a streamlined design with dedicated CPU and GPU airflow tunnels for seriously efficient heat dissipation. This advanced design provides better cooling for both the GPU and CPU, enabling heat to be expelled into the surrounding ambient air for greater energy efficiency and overall savings on system power. GPU and CPU Airflow tunnel


2+2 Power Supplies

Up to 96% efficient for reduced operating costs

Supports up to four 3000 W 80 PLUS® Titanium redundant power supplies to reduce operating costs and enable easier servicing. When four power supplies are installed, the configuration can be set to redundant with 2+2 or 2+1 models. This ensures power is supplied to the system equally from all power supplies to maximize efficiency, enabling uninterrupted operation.


Storage

Scalable design for growing workloads

designed for easy scalability to fit growing businesses and workloads. With PCIe Gen 5.0 x16 slots that fit full-height, full-length GPU cards, ESC8000-E11 features different combinations of storage and networking expansions and multiple interface types for flexible storage configurations.


3 PCIe

2 PCIe + 2 NVMe

1 PCIe + 2 NVMe + 1 OCP 3.0

Management

Comprehensive IT infrastructure solutions

Advanced ASUS management tools comprise IT assets management, security and reliability to improve the performance, up-time and real-time monitoring of the servers remotely — highly increasing the efficiency of IT infrastructure.


BMC


Remote server management

ASUS ASMB11-iKVM is the latest server-management solution from ASUS, built upon the ASPEED 2600 chipset running on the latest AMI MegaRAC SP-X. The module provides various interfaces to enable out-of-band server management through WebGUI, Intelligent Platform Management Interface (IPMI) and Redfish® API.


IT infrastructure-management software